Chromatin is a complex of histone proteins and genomic DNA, and serves to transduce the impact of environmental events on the transcriptome by either restricting or permitting the attachment of DNA binding proteins to the regulatory sequences of a gene. The H3K9 histone bimodal switch is pivotal in assembling either 'restrictive' (DimH3K9) or 'permissive' (AceH3K9) chromatin to gene-rich regions of the genome. The central hypothesis for this proposal is that, in schizophrenia, chromatin equilibrates towards a more restrictive state. This has direct clinical consequences because genomic DNA sequestered with restrictive chromatin, is inefficiently transcribed, may explain the less than optimal clinical response to synaptic psychopharmacology, and also because histone covalent modifications can be targeted with small molecule pharmacology. The specific aims in-toto will investigate the hypotheses: a) the H3K9 switch is differentially in the restrictive state (DimH3K9) in schizophrenia patients in both blood and brain; b) measurement of peripheral 'blood levels' of H3K9 in living patients will identify a homogeneous sample with lower levels of the permissive state (AceH3K9) that will respond preferentially to a clinical HDAC inhibitor such as valproic acid. Chromatin structure and function will be examined in both, postmortem brain samples from the Harvard brain collection, and circulating blood mononuclear cells (PBMC) from living patients obtained from three diagnostic groups; i) normals, ii) schizophrenia (both first episode and chronic), iii) bipolar disorder. In the postmortem brain, we will apply genome-wide molecular techniques (ChIP-seq) to survey and analyze the distribution of DimH3K9. Data from the ChIP-seq experiments on postmortem brain will reveal gene networks that are entrained by this restrictive chromatin and have repressed mRNA transcription. Accordingly, mRNA output from DimH3K9 repressed networks will be verified and compared across diagnostic groups using qRT-PCR. In living patients, we will quantify 'blood levels' of both DimH3K9 and AceH3K9 from nuclear extracts of blood mononuclear cells; those subjects with the lowest levels of AceH3K9 (and thereby possessing higher levels of restrictive chromatin) will be randomized to a controlled trial with the HDAC inhibitor valproic acid (with gabapentin as anticonvulsant control) to explore the therapeutic potential of medications capable of 'relaxing' chromatin. At the end of this project, we would have conducted a survey of chromatin assemblies in two major mental disorders and be positioned to answer several questions relating to epigenetic gene regulation in a clinical population. Equally important we begin the translation of published findings from cell, animal and post-mortem brain investigations into living clinical populations in an area with great theoretical and therapeutic implications.